The influencing factors of the erythropoietin resistance index and its association with all-cause mortality in maintenance hemodialysis patients

Abstract

Anemia is a common complication of chronic kidney disease with major option treatment of erythropoiesis-stimulating agents (ESAs). This study aimed to investigate the influencing factors of erythropoietin resistance index (ERI) and its association with mortality in maintenance hemodialysis (MHD) patients. Patients enrolled from China Dialysis Outcomes and Practice Patterns Study (DOPPS) 5 were included. ERI was calculated as follows: ESA (IU/week)/weight (kg, post-dialysis)/hemoglobin level (g/dL). The Cox regression model was used to analyze the influencing factors on survival outcomes. Stepwise multivariate logistic regression was used to identify the related risk factors, and subgroup analyses were performed. A total of 1270 MHD subjects (687 males and 583 females) were included, with an average age of 60 (49.0, 71.0) years. All subjects were divided into two groups by the median ERI of 14.03. Multivariate logistic regression showed that dialysis vintage (OR 0.957, 95% CI: 0.929-0.986), white blood cells (OR 0.900, 95% CI: 0.844-0.960), high flux dialyzer use (OR 0.866, 95% CI: 0.755-0.993), body mass index (OR 0.860, 95% CI: 0.828-0.892), males (OR 0.708, 95% CI: 0.625-0.801), and albumin (OR 0.512, 95% CI: 0.389-0.673) had a negative association with high ERI baseline (all p < 0.05). There were 176 (13.9%) deaths in total including 89 cardiac/vascular deaths during follow-up. Cox regression analysis showed that ERI was positively associated with all-cause mortality, especially in some subgroups. ERI was associated with increased all-cause mortality in MHD patients, indicating the possibility of death prediction by ERI. Patients with high ERI warrant more attention.

Keywords: Anemia; erythropoietin resistance index; maintenance hemodialysis; mortality.

Figure 1.

The proportional hazard regression (PHREG) curve of survival probability between high ERI and low ERI group. The shadow area was the 95% confidence interval. (a) all-cause mortality; (b) cardiovascular mortality.

Figure 2.

Cox regression analysis for ERI and all-cause mortality (a) or CV mortality (b). Model 1 was a crude model. Model 2: model 1 + age, gender, BMI, dialysis vintage, primary kidney disease. Model 3: model 2 + modified Charleson Comorbidity Index, fistula use. Model 4: model 3 + albumin (Alb), standardized kt/v.

Figure 3.

Subgroup analyses by gender (male vs. female), age (<60 vs. ≥60 years), diabetes, stdkt/v (<2 vs. ≥2), type of vascular access (fistula vs. catheter), time of prescribed dialysis per week (<3 vs. ≥3), high flux dialyzer, albumin (<4 vs. ≥4 g/dL), urine output (<200 vs. ≥200 mL/day), and congestive heart failure using cox regression. The model was adjusted for age, gender, BMI, vintage, and primary kidney disease.